It has seemed to us that the test of instruction might bear upon three exercises, which are easily marked--reading, arithmetic, and spelling. Here is a very simple table of tests (p. 54), of which we have made much use. It has been arranged with the help of M. Vaney. The table is suited to the elementary and to the intermediate course, and that is sufficient for examining defectives, since none of them are found in the senior division. It is scarcely necessary to say that this table of tests is the outcome of careful experiment. We have established for each age the average acquirements of all the children of that age whatever their place in school. One might quite as well have taken into account only the results given by typical children in the class proper to their age, but on reflection we rejected this proceeding as arbitrary, because it is affected by the difficulty of the curriculum, which is constructed _a priori_, whilst the average furnished by all the children of a given age is less artificial and is an adequate expression of the reality. Let us remark in passing that these two methods of calculation do not lead to equivalent results. The average furnished by the _typical_ children is higher than that furnished by _all_ the children, for, as we have shown above, more children are backward than in advance. Lastly, the time of year when the tests are made is not a matter of indifference. For spelling and arithmetic the time chosen was the end of February--that is, the middle of the session. For reading we are obliged to make use of results a little more advanced, for they were furnished later, namely, in June. SCALE SHOWING KNOWLEDGE ACQUIRED BY PUPILS OF ELEMENTARY SCHOOLS.